Objective
To investigate the influences of clinicopathological factors on the prognosis of patients with single-organ oligometastasis of oesophageal cancer (soOMEC) after surgery and to establish prognostic nomograms.
Methods
This was a retrospective analysis of 144 patients with soOMEC after surgery in a single centre who met the inclusion criteria. First, Cox univariate and multivariate models were used to analyse (SPSS 25.0 statistical software) the characteristics of patients, and independent prognostic factors for postoperative overall survival (OS) and OS after oligometastasis (OM-OS) were determined. Prognosis was analysed using R language software, nomograms were created based on the Cox multivariate analysis results, a bootstrap method (b = 200) was used for internal validation, and receiver operating characteristic (ROC) and calibration curves were used to validate the models.
Results
From January 2014 to December 2017, a total of 1595 patients with oesophageal cancer received R0 resection. As of the end of the follow-up period, 144 patients had single-organ oligometastasis (soOM). The median time to oligometastasis (TTO) in the whole group of patients was 14.2 months, and the 1-, 3-, and 5-year OS rates were 75.7%, 28.2%, and 13.3%, respectively. The median OS was 25.0 months (95% confidence interval (CI): 21.8–28.2); the 1-, 2-, and 3-year OS rates after distant metastasis (DM-OS) were 25.5%, 13.3%, and 7.2%, respectively, and the median DM-OS was 5.5 months (95% CI: 3.9–7.1). The Cox multivariate analysis results showed that three indicators, i.e., TNM stage (hazard ratio (HR) = 2.192, 95% CI: 1.441–3.336, P = 0.000), TTO (HR = 0.119, 95% CI: 0.073–0.194, P = 0.000), and treatment after DM (HR = 0.784, 95% CI: 0.970 − 0.025, P = 0.025) were independent prognostic factors affecting the OS of patients; TTO (HR = 0.669, 95% CI: 0.455–0.984, P = 0.041) and treatment after DM (HR = 0.713, 95% CI: 0.559–0.910, P = 0.007) were independent prognostic factors affecting the DM-OS of patients. Using the Cox multivariate analysis results, prediction nomograms for total OS and DM-OS of patients were established. In the validation of the nomogram models, the areas under the curve (AUCs) for the 1-, 3-, and 5-year total OS were 0.930, 0.927, and 0.928 in the training set and 0.705, 0.856, and 1 in the validation set, respectively; the AUCs for the 1-, 2-, and 3-year DM-OS were 0.904, 0.923, and 0.908 in the training set and 0.928, 0.842, and 0.895 in the validation set, respectively. The results showed that the two models have strong discriminative ability and good clinical promotion and application value.
Conclusions
Aggressive local therapy combined with systemic chemotherapy can benefit patients with soOMEC after surgery, and for patients with OM appearing at 1 year after surgery, aggressive radiotherapy or combined chemotherapy is expected to improve the prognosis and prolong OS. The nomogram models developed in this study are effective clinical tools for predicting the prognosis of such patients and can be of great value in predicting the prognosis of and determining the treatments for patients and can guide the individualised treatment of such patients.